Overview
Interactive charts can help to tell stories more effectively, by allowing the viewer to interact and do experiential learning. It helps the user to engage intellectually, take initiative and make their own visualisations based on their understanding and curiousity. A single static chart at best shows answers to a few questions, but users might want to develop specific insights or go deeper into the data to understand relationships between the data which cannot be achieved in a single chart.
We have done a static chart in the previous Take-home Exercise 1, where we went through the process of data import, data wrangling, combining, plotting the data, and refining the chart further using themes, and we produced a final output as shown below:
The above chart, while simple, does not give us a lot of information.
Firstly, it does not give us context. The chart only shows the Population of Singapore in 2021, but we do not know how the shape of the population pyramid has changed over the years. As mentioned previously, assessing a population pyramid over time can determine the stage of demographic transition
It also does not provide a detailed breakdown of the data. The data obtained from the Department of Statistics, Singapore, breaks down the data into Residents by Planning Area/Subzone, Age Group, Sex and Type of Dwelling. For a static chart, we can only show a combined version of the data and hence grouped everything and show the Age Group and Sex distributions.
Getting Started
In this exercise, we will create an one interactive chart allowing the user to decide what information they want to see, and another animated chart to show the changes in population over time.
Sketch of Design
First, we plan what we want to show on the graph. This allows us to know what output we want to achieve, and hence what data we need, how to wrangle the data and plot the graph.
Loading the required libraries
The code chunk below is used to check if the necessary R packages are installed in R, and if not, they will be. If they are already installed, they will be launched in the R environment.
packages = c('tidyverse', 'knitr', 'lemon',
'ggiraph', 'plotly', 'gganimate')
for (p in packages){
if(!require(p, character.only = T)){
install.packages(p)
}
library(p, character.only = T)
}
The knitr package is not part of the core tidyverse package and has to be listed separately. The lemon package helps to further customise the legends and axes and acts as a supplement to ggplot2.
Data Import
Like the previous exercise, the Population Data from Department of Statistics, Singapore is obtained.
We will navigate to the Population page and download both the Singapore Residents by Planning Area / Subzone, Age Group, Sex and Type of Dwelling, June 2000-2010 and Singapore Residents by Planning Area / Subzone, Age Group, Sex and Type of Dwelling, June 2011-2020 datasets.
As the data is provided in csv, we have to use the read_csv function which is already included in tidyverse to import data.
popn_0010 <- read_csv('data/respopagesextod2000to2010.csv')
popn_1120 <- read_csv('data/respopagesextod2011to2020.csv')
We now have 2 tibbles loaded into the Environment as shown.
Data Wrangling
We are interested in the total population of each Age Group for each of the Sexes for each Planning Area. Hence the filter, group_by, summarise and mutate of dplyr are used.
options(scipen=999)
popn <- bind_rows(popn_0010, popn_1120)
total_popn <- popn %>%
group_by(AG,Time, Sex) %>%
summarise(Population=sum(Pop)) %>%
ungroup()
total_popn$AG <- gsub("_"," ", total_popn$AG)
total_popn$AG <- gsub("\\<0 to 4\\>","00 to 04", total_popn$AG)
total_popn$AG <- gsub("\\<5 to 9\\>","05 to 09",total_popn$AG)
The option(scipen=999) is set to ensure that scientific notation will not be used in the charts.
Firstly, we combine both tibbles together using the bind_rows function.
We then group the tibbles by Age Group (AG), Time, and Sex, and summarise the statistic Population (Pop) by summing it. We then ungroup it to ensure that the data will not produce errors in future calculations.
Lastly, the data is made neater when the underscores in the Age Groups are replaced with spaces. We also replace the 0 to 4 and 5 to 9 age groups to ensure that they are sorted when plotting.
This will give us a tibble as seen below.
Interactive Chart
We can now create a basic interactive chart that shows the population of Males/Females in Singapore in a certain year based on the age groups.
Plotting the Data
First, we assign a tooltip to what we want to show on text hover, in this case we want to show the Population of an Age Group in a specific Year.
name <- paste0("\n", total_popn$Sex,
" Population of Ages ", total_popn$AG,
"\nin the Year ", total_popn$Time,
"\n= ", total_popn$Population)
We then use the same code as in Take-home Exercise 1, while changing the Title and Subtitle to ensure clarity. We also add a label = name in the mapping to show the tooltip later.
In this plot, we also pass through a colour through the geom_col function to clearly show the separation of Years.
p <- ggplot(data = total_popn,
mapping = aes(
x=ifelse(test = Sex=="Males", yes=-Population,
no=Population),
y=AG,
fill=Sex,
label = name)) +
theme(panel.background = element_rect(fill="white"),
panel.border = element_rect(linetype=1, fill=NA)) +
geom_col(color="black") +
labs(x="Cumulative Population") +
labs(y="Age Groups") +
labs(title = "Population of Singapore",
subtitle = "Aggregated, 2000-2020") +
scale_x_symmetric(labels=abs)
p
Next, we use the ggplotly function of the plotly package and change the default tooltip (which shows every mapped value) to the tooltip we want with the code chunk below.
ggplotly(p, tooltip = c("label"))
We can now interact with the chart, and it shows the data that we want to show on hover, great! Other interactive elements include the ability to filter Male and Female population only. We can also zoom in on the chart if there is a particular data point that we are interested in, as shown below.
Animated Chart
Next, we want an animated chart to show the changes in the population pyramid structure over time.
Plotting the data
We can make use of the earlier ggplot object, p, and change the labels including the x-axis, Title and remove the subtitle to prepare for our animated plot with the code chunk below.
We use a parameter frame_time as an input to show the year when we animate it later.
p2 <- p +
labs(x="Population") +
labs(title = "Population of Singapore,
{as.integer(frame_time)}", subtitle = NULL)
p2
Do not worry about the population values as they will be sub-divided using the Time column.
Lastly, we will use the transition_time and ease_aes functions of the gganimate package to show how the data changes over time, and pass the Time column (which shows the years) as an animation.
panim <- p2 +
transition_time(Time) +
ease_aes('linear')
panim
Conclusion
With these plots, we can see how the charts have become more interactive and animated, to show the user a clearer picture of the data and to allow the user to pinpoint what data he/she wants to find out.
Intepretation of results
For example, for the animated population pyramid, the policy-makers can see the changes in the population shape over time, and implement policies according to the changes based on the population structure.
We can see that the population is aging over time, with the base of the pyramid becoming smaller over time. This shows that either the fertility rate in Singapore is decreasing, or the mortality rates of young people in Singapore is increasing. Since we have local context, we know that the former is true and not the latter.
Next, we can see that the tip of the pyramid is expanding. This shows that there is a larger elderly population and that a higher age expectancy in Singapore. This could be due to a better healthcare system and elderly-care facilities.
Possibilities of extension
The exercise can be further extended of course, with the use of variable inputs and outputs, such as using R Shiny and Dash from Plotly. These two packages allow the user to customise what inputs they need, such as the Planning Area, Subzone, Age Groups, and visualise in the form of various outputs.